crow, and had found it to be gray. Why is a single instance in some cases sufficient for a complete induction, while in others myriads of concurring instances, without a single exception known or presumed, go such a very little way toward establishing a universal proposition? Whoever can answer this question knows more of the philosophy of logic than the wisest of the ancients, and has solved the great problem of induction."

Hamilton attempts a solution by propounding a distinction between characters that are essential and characters that are unessential. Still he adds, "the difference of essential and accidental is one itself founded on induction, and varies according to the greater or less perfection to which this has been carried. In the progress of science, the lateral lobes of the cerebellum may appear to future physiologists as necessary a condition of the function of suckling their young as the organs of breathing appear to us of circulation and of life."

Hamilton's distinction of essential and accidental characters seems to involve the same theory that is presented by Professor H. N. Day,2 in which all inductive reasoning is made to depend upon the relation of part to complementary part. From one part of a causal whole, all the parts are necessarily inferred. There seems, however, to 1 Logic, p. 453. 2 Logic, pp. 195–202.

be some doubt whether causal is used by him with reference to "final cause" at all, or wholly with regard to secondary causes.

President Porter,1 advocating this distinction between essential and accidental characters, as giving validity to inductive reasoning, would determine what is essential and what accidental by a judicious appeal to the intuition of final causes. In his opinion, the decisive reason why we believe it to be a uniform law of nature that men's heads are situated above their shoulders on so much less evidence than we would believe it a uniform law of nature that birds having the anatomical structure of swans must always be white, is that the position of the head in the anatomical structure is an essential attribute, while the color of the skin is an unessential attribute. The position of men's heads in their anatomical structure is essential, because "otherwise they could not perform the functions of men with any convenience or success. Such a form would offend both the eye and the mind, and would be entirely incompatible with the ideal of beauty and convenience to which we assume nature would certainly conform." "Considerations of convenience and of adaptation, and even of beauty and grace," he adds, "go far towards deciding the question. They give that weight and force to those 'single instances

1 Human Intellect, pp. 481, 482, et al.

which in some cases are sufficient for a complete induction,' and detract all force from the myriads of concurring instances' in other directions." The assumption is here made that we are created in the image of God, and hence can positively interpret his ends and methods in creation. The other assumptions which he considers as underlying the inductive process are the reality, "first, of the distinction of substance and attribute; second, of the causative relation; third, of time and space; fourth, of uniformity in the indications and operations of nature; fifth, of the adaptation of the beings and powers of nature to certain ends." Upon these he asserts, "the entire process of induction rests, and upon their validity is founded its trustworthiness." The object of the present discussion is to consider more critically and fully the degree of validity belonging to these various assumptions, and their relation to the real basis of confidence in induction.

II. 1. In regard to the assumptions that there is a distinction between substance and attributebeing and phenomena - and, what amounts to the same thing in the end, the assumption that there is a relation of causality in the sense of secondary, or efficient, causality, we remark that these may be ruled out of the case, as tending to entangle the real problem with useless questions concerning ontology, or the nature of things in themselves.

We need not for the purposes of the present discussion concern ourselves with the question whether we can by experience get back of the attributes of substance, the phenomena of being, or the effects of causes. So far as all confidence in inductive reasoning is concerned, we shall not be embarrassed by adopting the phrase of Mill, permanent possibility of phenomena," as being all there is between us and God. The words "permanent possibility" would here cover all the ground of secondary causes needed for the argument.

66

2. The assumption with regard to the so-called law of uniformity in nature needs restatement.

What, for example, do writers mean when they say that belief in the uniformity of nature's operations is the ground of our confidence that the sun will rise to-morrow? They certainly do not speak of an absolute uniformity, for no one denies that the movements which occasion the phenomena of sun-rising had a beginning. The nebular hypothesis agrees with others regarding that. Nor do they mean that these motions are necessarily to continue forever. This appears forcibly in the account Mill gives of his belief in to-morrow's sun-rising.1 He believes that he discovers so many causes in operation to produce the rising of the sun to-morrow, that there could not be coun

1 Logic, Book iii. chap. 19.

teracting causes enough in operation to prevent it in so short a time, without exhibiting signs of their presence to-day. Still he admits that, if asked whether the sun will rise a million years hence, he has no answer.

As it stands this answer of Mill is evidently fallacious. What ground has he for believing that causes sufficient to counteract the forces producing the phenomena of the sun-rising to-morrow, must exhibit themselves to-day, i.e. one day before their effectuation? Why does a million of years change the conditions of the problem? Has he never heard of nitro-glycerine, which presents a constant and harmless set of phenomena under a great variety of circumstances, but only awaits some otherwise insignificant changes to complete the conditions necessary to bring its explosive properties into working order? What grounds of confidence have we that there are not in the solar system forces analogous to this? If it is answered: such phenomena are extremely rare in nature's large operations, we ask what is a long interval in eternal time, and what is a large operation? Our experience, for good reasons, has not taught us the exact accompaniments of the explosion of a solar system; but certainly there is no natural impossibility in its taking place without premonition. Nor is there any allowance made in this answer for the supposition that a personal